Water scarcity, possibly caused by global climate change or climatic variability and global population growth, has led to increased difficulties in the utilization and management of water resources [1
]. Among the various water resources, groundwater resources are usually considered to be relatively less susceptible to climate change [4
]. However, population growth and socioeconomic development have led to a continuous increase in the need for groundwater resources. During dry seasons or in arid regions, groundwater is the main source of water and a key factor in discharge control. This underlines the importance of enhanced planning, the implementation of water resource management and allocation policies, and a good understanding of the components and characteristics of the catchment hydrological cycle, which facilitates the sustainability assessment of groundwater exploitation and management. However, in remote or larger study areas, the use of relevant technologies and the execution of surveys are usually associated with a higher cost. Transmissivity, hydraulic conductivity, specific yield, and effective aquifer depth are the basic parameters that describe groundwater hydrology as well as the necessary input data for many physical models. In the past, the values for these hydrogeological parameters were mainly acquired through local-scale field or laboratory tests. Therefore, the hydrogeological structure of larger areas or areas containing sloping aquifers could not be effectively determined. In addition, because of the high cost and complexity of investigating catchment-scale hydrogeological parameters, there has been a lack of research and discussion on this issue.
Among the estimation methods for catchment-scale hydrogeological characteristics, flow recession analysis is an effective technique with a relatively high developmental potential [7
]. This method has been commonly used in previous studies that have investigated the relationships between groundwater storage and streamflow. As streamflow is mainly composed of direct runoff, produced by rainfall, and baseflow, contributed by groundwater storage, only baseflow needs to be considered in the investigation of storage-discharge relationships. However, as the baseflow may include lateral flow caused by rainfall infiltration, it is not an effective representation of the baseflow contributed solely by groundwater. Therefore, the World Meteorological Organization (WMO) has defined low flow as the “streamflow during a prolonged dry period,” i.e., the streamflow sustained by the natural discharge of groundwater from upstream aquifers during the no-rainfall period [11
]. On the basis of the low-flow concept, Brutsaert and Nieber [12
] utilized the Boussinesq hydrological model, for groundwater discharge from an unconfined aquifer into a channel, and the hydraulic groundwater theory to propose a low-flow recession analysis method. Based on the fitting of streamflow data, this method can estimate the catchment-scale discharge characteristic constants and be used in studies investigating groundwater storage in catchments, spatial distribution of aquifer properties, and channel networks with landscape features [13
]. This method has been widely applied in studies with aims including the estimation of hydraulic characteristics of groundwater [14
], trends in groundwater storage [16
], drought detection [18
], and regional low-flow analysis [20
]. Another method that can be used in the estimation of hydrogeological parameters is the recession-curve-displacement method developed by Rorabaugh [22
]. This method is based on the recharge theory, which describes the relationship between rainfall infiltration and stream level variations. It involves the estimation of transmissivity through streamflow hydrographs. It has been applied in previous studies estimating groundwater recharge and transmissivity [23
], which indicates its practical applicability in different regions and environments. In another study, Rutledge [24
] developed the RECESS and RORA programs for automation of the recession-curve-displacement method in order to reduce the subjectivity inherent in manual methods and to provide enhanced reliability and usability. These two methods mainly involve the analysis of hydrological data that can be readily acquired. The streamflow during no-rainfall periods represents the overall discharge behavior in the upstream catchment area of the gauge station. Compared to the local-scale measurements, these methods enable the determination of overall hydrogeological characteristics in the catchments and investigation of the hydrological process from a macro perspective [6
During recent years, climate change has adversely affected Taiwan, with the effects manifesting as an aggravated and increased frequency between wet and dry seasons, increased temperatures, an increased number of extreme events, a decreased number of rainy days, and the increased intensity of rainfall. As a result, the hydrological uncertainties of catchments in various region of Taiwan have increased. Combined with the inherent, uneven spatiotemporal distribution of rainfall and ineffective runoff retention because of steep slopes and high streamflow velocities, this has led to an increasingly severe water supply crisis. In Southern Taiwan, where the effects of climate change are particularly serious, the rainfall ratio between wet and dry seasons is 9:1. During the wet season, rainfall manifests in typhoon and flood periods. The high-intensity rainfall scours large amounts of sediment, which increases river turbidity and reservoir sedimentation, thereby reducing the effective storage capacities of the reservoirs. During dry season, water shortages are likely to occur because of the substantial reduction in streamflow. In order to alleviate water shortages, this leads to an increased dependence on existing water resources for agriculture and river ecosystems [26
], which may impact ecological habitats in the upstream area. The Water Resources Agency [27
] has also reported the significant trend of declining streamflow in Southern Taiwan under the most adverse, simulated, future rainfall scenario conditions, which further emphasizes the criticality of future water resource issues in Southern Taiwan. Therefore, there is a need to enhance water resource regulation in the region under current climate change conditions.
In view of the aforementioned situations, the present study utilized a combination of two flow recession analysis methods for the analysis of data from nine hydrological gauge stations in Southern Taiwan. The aim was to estimate the discharge characteristics of catchments and catchment-scale hydrogeological parameters in the region and to investigate the influence of seasonality on the estimation results. Subsequently, the estimated hydrogeological parameters were compared to previous field test results to determine the applicability of the methods used in this study and whether the estimated results showed significant regional differences and consistency with the distribution of the geological structures. The results may be used as a reference for parameter setting in hydrological models and for future planning and decision-making in water resources management.
2. Study Area
Southern Taiwan consists of several administrative divisions, including Chiayi County, Chiayi City, Tainan City, Kaohsiung City, and Pingtung County. The total area of the region is 10,002 km2
, accounting for approximately 28% of the total area of Taiwan. Because of its tropical monsoonal climate, terrain, and geographical location, the rainy season is influenced by southwestern monsoons and typhoons from May to October. The dry season is influenced by northeastern monsoons from November to April and the leeward side causes less rainfall in this region [26
]. The main rivers in the study area are the Bazhang, Zengwun, Yanshui, Erren, Kaoping, Donggang, and Linbian. In the present study, streamflow data from nine gauge stations in the catchments of Southern Taiwan were classified into dry season (November–April) and wet season (May–October) for the estimation of the respective catchment discharge characteristics and hydrogeological parameters during the dry and wet seasons. Table 1
shows information on the gauge stations in the catchments of Southern Taiwan.
Regarding the geological conditions of the study area, the Chianan Plain is primarily characterized by littoral and lagoon faces. Apart from a small number of areas with channel-fill deposits, there are no large fluvial deposits in the plain. On the North Chianan Plain, because of its slow sedimentation rates and longer emergent periods, there is an obviously weathered soil layer, and the relatively fine-grained sand accounts for a smaller proportion. On the South Chianan Plain, tectogenesis has divided the region into several separate underground zones by the consolidation of marine argillaceous strata. Although the sand layer accounts for a higher proportion of strata, the thicknesses of the sandy strata vary greatly because of tectonic influences. In addition, the widely distributed partially consolidated marine mudstone also reduces the lateral connectivity in the strata. These stratum characteristics result in low groundwater recharge from rainfall infiltration and, thereby, a shortage of groundwater resources on the Chianan Plain [28
]. On the Pingtung Plain, metamorphism increases from west to east and the strata of the foothills, mainly consisting of black slate mixed with quartzite, contribute to the low permeability. The main aquifers within the plain are composed of unconsolidated rock, including older conglomerate alluvial fan deposits and, the most widely distributed, recent alluvium (Figure 1
). The aquifers of the Pingtung Plain are well developed with relatively large thicknesses and a wide distribution. The thickness of the near-surface unconfined gravel aquifers decreases towards the west, with aquifuge thickness being much less than aquifer thickness. Within the plain, a continuous distribution of aquifuge only exists in the coastal areas [29
]. According to reports of the differences in the spatial distribution of geology, we divided Southern Taiwan into the Chianan (Zengwun, Yanshui, and Erren river basins) and Kaoping (Kaoping, Donggang, and Linbian river basins) sub-areas in our discussion of the hydrogeology parameters.
In the present study, a combination of the low-flow recession analysis method, which is based on the water balance in hydrological systems, and the recession-curve-displacement method, which is based on streamflow hydrographs, was used for the estimation of hydrogeological parameters of catchments in Southern Taiwan. The recession index K derived from the low-flow recession analysis method was used to determine the seasonal differences in the discharge regimes of aquifers in the catchments. Subsequently, the respective estimation results for transmissivity T, hydraulic conductivity k, and specific yield Sy for the dry and wet seasons were compared to the results of field tests. The comparison results indicated small differences between the estimation results for the dry and wet seasons as well as slight differences in the local-scale and catchment-scale hydrogeological parameters. The graphs of hydraulic conductivity vs. specific yield showed significant regional differences and consistency with the distribution of geological structures. They demonstrated the applicability and representativeness of the low-flow recession analysis method in the estimation of hydrogeological parameters. The present study differs from costly field tests, which merely produce local-scale results for the estimation of hydrogeological parameters, because hydrological data that could be readily acquired and were representative of catchment discharge behavior were used. A combination of two flow recession analysis methods was employed to eliminate subjectivity in the initial estimation of one parameter, a problem that has frequently been encountered in previous studies. The estimated catchment-scale hydrogeological parameters demonstrated that flow recession analysis can provide a rapid, low-cost, and effective means of estimating hydrogeological parameters, which can facilitate the construction of future hydrological models and a better understanding of the role of groundwater in various catchments or basins of Taiwan in the hydrological cycle. They can also serve as a reference for decisions regarding groundwater resource management under conditions of climate change.